Peptide-Drug Conjugate Targeting Keratin 1 Inhibits Triple-Negative Breast Cancer in Mice

Abstract

Selective delivery of chemotherapy to the tumor site while sparing healthy cells and tissues is an attractive approach for cancer treatment. Carriers such as peptides can facilitate selective tumor targeting and payload delivery. Peptides with specific affinity for the overexpressed cell-surface receptors in cancer cells are conjugated to chemotherapy to afford peptide-drug conjugates (PDCs) that show selective uptake by cancer cells. Using a 10-mer linear peptide (WxEAAYQrFL) called 18-4 that targets and binds breast cancer cells, we designed a peptide 18-4 – doxorubicin (Dox) conjugate with high specific toxicity toward triple-negative breast cancer (TNBC) MDA-MB-231 cells and 30-fold lower toxicity to normal breast MCF10A epithelial cells. Here we elucidate the in vivo activity of this potent and tumor-selective peptide 18-4 – Dox conjugate in mice bearing orthotopic MDA-MB-231 tumors. Mice treated with four weekly injections of the conjugate showed significantly lower tumor volumes compared to mice treated with free Dox at an equivalent Dox dose. Immunohistochemical (IHC) analysis of mice tissues revealed that treatment with a low dose of PDC (2.5 mg/kg of Dox equivalent) reduced the expression of proliferation markers (PCNA and Ki-67) and increased apoptosis (evidenced by increased caspase-3 expression). At the same dose of free Dox (2.5 mg/kg), the expression of these markers was similar to that of saline treatment. Accordingly, significantly more Dox accumulated in tumors of conjugate treated mice (7-fold) compared to the Dox treated mice, while lower levels of Dox was observed in the liver, heart, and lungs of peptide-Dox conjugate treated mice (up to 3-fold less) than Dox treated mice. The IHC analysis of keratin 1 (K1), the receptor for peptide 18-4, revealed K1 upregulation in tumor and low levels in normal mammary fat pad and liver tissues from mice suggesting preferential uptake of PDCs by TNBC to be K1 receptor-mediated. Taken together, our data support the use of a PDC approach to deliver chemotherapy selectively to the TNBC to inhibit tumor growth.

Graphical Abstract

Tumor-selective peptide-drug conjugate (PDC) inhibits TNBC in mice

INTRODUCTION

Peptide-drug conjugates (PDCs) are emerging as a new modality for cancer treatment.^1-3 In a PDC (Figure 1A), a cytotoxic drug (payload) is conjugated to a targeting peptide to specifically deliver the drug to the tumor and/or tumor microenvironment, thereby potentiating the therapeutic activity of the drug. While PDCs are under development, antibody-drug conjugates (ADCs), where a monoclonal antibody is a targeting ligand, is now an established therapeutic approach for delivering cytotoxic drugs specifically to the cancer site.^4-5 The approach of improving the cancer specificity of payloads by conjugating them to a targeting ligand allows increasing the therapeutic index of highly cytotoxic to moderately cytotoxic drugs and their administration with less side effects.⁶ The targeting ligand, a peptide in a PDC or an antibody in an ADC, binds overexpressed cell-surface receptors in cancer cells for specific internalization of the conjugate. PDCs being small (~3-5 kDa), offer several advantages over large conjugates, like ADCs, made with antibodies (~150 kDa).^1-3 PDCs allow better tumor penetration, homogenous solutions with less batch-to-batch variation, easy production, low manufacturing costs, low or no immunogenicity, and a better ability to cross BBB potentially targeting cancer metastasis in the brain. Furthermore, peptides have reduced binding affinity to the target receptors compared to the monoclonal antibodies and, therefore may alleviate on-target non-tumor binding and associated toxicities when used at low concentrations as the target cell-surface receptors are not exclusively present on cancer cells.⁷

Figure 1.

A, Schematic showing a peptide-drug conjugate for targeting cell-surface receptors in cancer cells via a carrier peptide. B, Chemical structure of peptide 18-4 – doxorubicin (Dox) conjugate used in this study. Peptide, linker, and drug are shown in red, blue, and black, respectively. Norleucine (x) and arginine (r) are D-amino acids.

Drug conjugates target cell-surface receptors (CSRs) that show higher expression in cancer cells compared to normal cells. Cancer cells should display at least 3-fold higher expression of a CSR compared to the normal cells, and the amount of CSR overexpressed by cancer cells should be sufficiently high to ensure an adequate amount of drug is delivered to increase the therapeutic index of the targeted drug.² Cell-surface keratin 1 (K1) is a novel target for solid cancers of epithelial origin.⁸ K1 is differentially expressed in cancer and normal cells.⁸ K1 is a type II intermediate filament protein that plays an important role in cell growth and proliferation, and aberrant regulation of keratin expression is linked to human diseases.^8-10 K1 is a cytoplasmic protein; however, cell transformation events like tumorigenesis and apoptosis lead to increased and cell-surface expression of K1. Cancers like breast, neuroblastoma, nasopharyngeal, and hepatocellular carcinomas show increased cell-surface K1 expression and/or overexpression of K1 compared to normal cells and tissues.⁸ The differential and cell-surface expression of K1 in cancer over normal cells makes K1 an attractive target for delivering drugs specifically to cancer cells.

We engineered peptides that bind to cell-surface K1 and are internalized by breast cancer cells via cell-surface K1 receptor-mediated endocytosis.^11-12 Peptide 18-4 (WxEAAYQrFL), with two D-amino acids, is a second-generation breast cancer cell-targeting peptide that is proteolytically stable (100% intact up to 24 h when incubated with human serum or liver homogenate from mice) and has shown high specific uptake by breast cancer cells and minimal/no binding to non-cancerous cells.¹² Affinity purification of breast cancer cell lysates using immobilized peptide, followed by liquid chromatography-tandem mass spectrometry and proteomics were used to identify K1 as the novel target for peptide 18-4 in cancer cells.¹¹ Further, we showed that the uptake of the peptide by the cancer cells is dependent on K1 expression.¹¹

Several PDCs have been prepared with peptide 18-4 and doxorubicin (Dox) using different linker chemistries such as ester, amide, succinimidyl thioether, or acyl hydrazone.^13-14 In vitro results showed that these PDCs were highly specific toward the breast cancer cells. PDCs displayed similar toxicity as free Dox toward the breast cancer cells and several-fold (7-40 times) less toxicity toward the non-cancerous cells such as MCF-10A and HUVECs. A peptide18-4 – Dox conjugate (Figure 1B) with amide/succinimidyl thioether linkage showed high selective toxicity toward triple negative breast cancer (TNBC) cell lines, i.e., MDA-MB-231 cells (IC₅₀ 1.3 ± 0.2 μM) and MDA-MB-468 cells (IC₅₀ 4.7 ± 0.3 μM) compared to the normal breast MCF10A cells (IC₅₀ 38.6 ± 1.1 μM).¹⁴ The linkage between the drug and the peptide was stable as the degradation half-life of peptide18-4 – Dox conjugate in the presence of human serum was found to be ~ 18 hours.¹⁴ Herein, we describe the first in vivo evidence for improved efficacy of this PDC targeting K1 receptor in an orthotopic TNBC mouse model. We also show a higher accumulation of PDC in TNBC tumors in mice, in accord with K1 overexpression in tumor over non-tumor tissues in MDA-MB-231 xenografted mice.

MATERIALS AND METHODS

Materials.

Chemicals and solvents were purchased from Sigma-Aldrich unless noted otherwise. The water (ultra-pure) used was from Milli-Q system. Purification and analysis using reversed phase (RP)-HPLC were conducted on a Prominence-i RP-HPLC system (Shimadzu Corp., USA) using C18 semi-preparative (10 mm × 250 mm, 5 μm) column. An autoflex speed MALDI-TOF mass spectrometer (Bruker, USA) was used for mass analysis. For analysis of mice tissue samples, Shimadzu LCMS 2020 system (Shimadzu Corp., USA) was used running with an analytical C18 Vydac column (4.6 mm x 250 mm, 5 μm) at a flow rate of 0.8 mL/min.

Peptide 18-4 – Dox Conjugate Synthesis and Purification.

The conjugate was synthesized in two steps. The first step involved reaction between Dox and sulfo-SMCC linker to give MCC-Dox as described before.¹⁴ The second step, that involved reaction of MCC-Dox and peptide, was modified (Figure 2). Briefly, peptide (1.1 mg, 0.78 μmol) in DMF (500 uL) was added to MCC-Dox (0.8 mg, 1 μmol, 500 uL DMF) solution, and PBS (500 uL, 100 mM, pH 7.4) was added. The reaction mixture was stirred at room temperature under nitrogen overnight. The reaction progress was monitored using RP-HPLC. The crude mixture of the conjugate obtained was purified using RP-HPLC as before.¹⁴ Pure conjugate was stored as a dry powder at −80 °C until use.

Figure 2.

Reaction between thiolated peptide 18-4 and MCC-Dox in DMF/PBS to give peptide 18-4 – Dox conjugate.

Cell Culture, Tumor Xenograft in Mice, and Treatment with the Conjugate.

The firefly luciferase-expressing human TNBC cell line MDA-MB-231 (clone D3H2LN) was received as a gift from the laboratory of Dr. Mary Hitt (University of Alberta), and the cell culture medium used was the following: MEM supplemented with 10% FBS, 1 mM sodium pyruvate, 0.1 mM NEAAs, 2 mM L-glutamine, 100 IU/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL Fungizone^®. MDA-MB-231-luc-D3H2LN cells were maintained at 37 °C in 5% CO₂ atmosphere. The cells were tested negative for mycoplasma contamination according to a PCR-based method using Mycoplasma Detection Kit (InvivoGen).¹⁵

Twenty-one female NIH-III mice (8-week-old) were purchased from Charles River (Wilmington, MA). The animal study was conducted in accordance with the guidelines of the Canadian Council on Animal Care (CCAC) with approval from the Animal Care and Use Committee (ACUC) of the University of Alberta (Edmonton, AB, Canada). An orthotopic breast cancer mouse model was generated by inoculating 2 × 10⁶ MDA-MB-231 cells into the left abdominal mammary fat pad (50 μL cell suspension composed of 50% PBS and 50% Matrigel Basement Membrane Matrix, BD Biosciences, Franklin Lakes, NJ). The health status of each mouse was assessed through body weight measurement and signs of distress (e.g., lethargy, ruffled coat, ataxia, etc.). Tumor growth was monitored by caliper and optical imaging. For measuring luminescence intensity, luciferin potassium salt solution (150 mg/kg) was intraperitoneally injected, and sequences of images were taken to determine the peak of luciferase activity.¹⁶ Animals were homogeneously allocated into 3 groups of 7 mice each, according to the tumor growth. The three treatment groups (n=7) were (i) 0.9% saline solution containing 6% DMSO and 6% tween 20, which was the diluent used throughout the groups, (ii) free Dox (2.5 mg/kg body weight), and (iii) peptide 18-4 – Dox conjugate (2.5 mg Dox equivalent/kg). All treatments were administered intravenously via tail vein, and the first day of injections was 10 days post cell inoculation, followed by the subsequent injections every 7 days, in a total of 4 intravenous (i.v.) bolus injections (Figure 3A).

Figure 3.

Peptide-Dox conjugate inhibits tumor growth in mice bearing orthotopic TNBC tumors. A, Schematic of the experimental design for efficacy study in mice. Eight-week-old female NIH-III mice (21 mice) were injected with MDA-MB-231 cells in mammary fat pads. Ten days later mice were randomly grouped into three groups (n =7) and treated once weekly i.v. dose of saline (group 1), free Dox (group 2), or conjugate (group 3). Mice were euthanized 24 h after the fourth treatment and tissues were dissected and collected for further analysis. B, Tumor growth in mice using calipers to obtain tumor volume. C, Tumor growth followed with optimal imaging (IVIS) of mice to obtain luminescence intensity. D, body weight of mice monitored during the study. E, representative optical images of mice from three treatment groups showing tumor bioluminescence. Each picture shows the capture of the firefly luciferase activity at two different timepoints: pre-treatment scenario (Day −3) and 24 h after the last injection (day 23).

In vivo Drug Biodistribution.

To analyze the biodistribution of peptide 18-4 – Dox conjugate and Dox, the mice were euthanized 24 h following the final dose. Tissues/organs (tumor, liver, heart, lungs, spleen, kidneys) were harvested from mice (n=3) from each group (Dox or PDC treated) and stored at −80 °C. To extract drug from each sample, tissue was thawed. Tissue (up to 200 mg) was extracted for conjugate/Dox analysis using LC/MS method as described previously.¹⁷ An internal standard, Daunorubicin (Dau, 150 μM, 20 μL) was added to the sample before extraction. As a control, liver from saline treated mice spiked with Dox (0.06-6 pmoles) and Dau was extracted and analyzed following the same method. In this concentration range, the LC/MS method detected measurable peaks (0.15 – 14.5 μg/g) from all extracted control samples.

Determination of Caspase-3, PCNA, and Ki67 in Mice Tumor Tissues.

Tissue samples collected from mice were also preserved and sectioned for immunohistochemistry (IHC) analysis. The dissected tissues from tumor from 3 mice of each group were collected. The tissues were immediately transferred to a neutral buffered 10% formalin solution for fixation and stored for at least 24 h at room temperature. This was followed by 1 hour in 70%, 95% and 100% ethanol and three days in 2-butanol. Then the samples were embedded in paraffin blocks and slides were prepared by slicing into 5 μm sections and mounting on glass slides. Tissue sections were deparaffinized and rehydrated. Heat-induced epitope retrieval was performed by boiling sections in citrate buffer (Sigma-Aldrich), pH 6.0, for 10 minutes. Sections were blocked with 1% fish gelatin followed by overnight incubation at 4°C with the following primary antibodies: caspase-3, Ki-67, and PCNA. Following blocking with 3% hydrogen peroxide, sections were incubated with the Dako EnVision+ System HRP Labelled Polymer secondary antibody (Agilent) for 1 hour at room temperature. The sections were developed using Dako DAB+ Chromagen (Agilent) and counterstained with hematoxylin.

Determination of Keratin 1 Expression in Mice Tissues.

For K1 expression, tumor and liver tissues were collected from mice bearing orthotopic TNBC tumors. In addition, tissues from the mammary fat pad were also collected from normal female NIH-III mice. The tissues were stained with rabbit monoclonal [EPR17744] to cytokeratin 1 (Abcam, USA), and similar steps were followed, as mentioned above to determine K1 expression in these tissue sections.

Statistical Analysis.

Graphs were prepared using GraphPad Prism8 (La Jolla, CA, USA). Data are presented as the mean ± SEM, and the statistical analyses were performed using the same software. Significant differences among groups were assessed using one-way ANOVA, followed by Tukey’s post-hoc test. The minimum level of significance was set for p < 0.05.

RESULTS

Peptide 18-4 – Dox Conjugate.

The peptide 18-4 – Dox conjugate (Figure 1B) was synthesized in two steps as reported previously.¹⁴ MCC-Dox was synthesized from Dox and sulfo-SMCC as before.¹⁴ The reaction between MCC-Dox and the thiolated peptide was modified to increase the conjugate yield. Previously, we reacted peptide with MCC-Dox in DMF with a catalytic amount of DIPEA for 4 hours to obtain conjugate in 63.5% yield.¹⁴ In the modified method, the peptide was reacted with MCC-Dox in DMF and PBS (100 mM, pH 7.4) to maintain basic pH (Figure 2), and the reaction was allowed to continue for 24 h at room temperature. These conditions gave a higher conjugate yield of 81%. The conjugate was purified and characterized using RP-HPLC and mass spectrometry (supporting information Figures S1 and S2). The purity was confirmed by RP-HPLC (>95% pure). The UV-vis spectrum for the pure conjugate was recorded that showed a maxima at ~481 nm (Figure S3).

The conjugate is hydrophobic as it elutes at 40% acetonitrile in RP-HPLC using a gradient elution method (30–45% acetonitrile/water containing 0.05% trifluoroacetic acid). The conjugate has limited aqueous solubility, not adequate for its solubilization at high concentrations needed (~ 0.5 mM) for intravenous (i.v.) injections in mice. Therefore, for i.v. injections, the conjugate was dissolved (dispersed) in normal saline with DMSO (6%) and tween 20 (6%). DMSO and tween 20 are commonly used to prepare solutions of lipophilic compounds for i.v. injections.¹⁸ Fresh solutions were made before i.v. injections in mice.

In vivo Antitumor Activity of the Conjugate.

To compare the activity of the conjugate with free Dox and saline, mice bearing orthotopic MDA-MB-231 tumors were divided into three groups (n=7) and treated once weekly with an i.v. injection of saline (control), Dox (control) or peptide-Dox conjugate for up to 3 weeks (Figure 3A). Mice were treated with a low dose of Dox (2.5 mg/kg) or conjugate (2.5 mg/kg Dox equivalent) for this study. Tumor growth in mice was measured using two methods, manual measurement using the calipers (Figure 3B) and optical imaging of the mice to record luminescence from the tumor (Figure 3C). While optical imaging is expected to be more accurate, both methods showed similar results. With just four treatments (day 1, 8, 15, and 22), the mice treated with the conjugate showed significant tumor growth inhibition relative to the control groups. Further, the mice body weights were not significantly affected between the treatment groups during the 23-day study period at the low Dox dose injected (Figure 3D). Figure 3E shows representative optical images of three mice from each group taken at pre-treatment (day −3) and post treatment (day 23) displaying intensities of the tumor luminescence.

Conjugate and Dox Biodistribution in Mice Tissues.

Organic extraction was used to separate peptide 18-4 – Dox conjugate or its metabolite Dox from different tissue samples collected from drug (peptide–Dox or Dox) treated NIH III mice. The extracts were analyzed using LC/MS to quantify the peptide-Dox conjugate or Dox using daunorubicin as an internal standard. Among all samples, the highest concentration of Dox was found in tumor tissues (0.22 ± 0.002 μg/g) from the conjugate-treated mice, and this was significantly higher (7x more) than in tumor from free Dox treated mice (0.03 ± 0.001 μg/g) (Figure 4). Next, Dox was also found in the liver, heart, and lungs. Liver tissues from the conjugate-treated mice showed Dox at 0.04 ± 0.008 μg/g, which was at significantly lower levels (3x less) compared to the Dox found in liver (0.12 ± 0.01 μg/g) from Dox-treated mice. Similarly, the heart and lungs showed lower levels of Dox in conjugate-treated mice than the free Dox-treated mice. Overall, the results show that the conjugate is selective for tumor with significantly high levels of Dox in tumor and low levels of Dox in other non-tumor organs compared to what is observed with free Dox-treated mice (Figure 4). Further, it is noted that the Dox levels detected in tumor tissues from NIH III mice (peptide-Dox conjugate treated) were ~8 times lower than what was found previously in NOD-SCID mice carrying MDA-MB-231 xenografts and treated similarly with Dox or peptide-Dox (at 2.5 mg/kg Dox or Dox equivalent).¹⁷ The reason for this observation is not clear and needs further investigation.

Figure 4.

Dox biodistribution in tumor and other tissues (liver, heart, and lungs) from mice at 24 h post treatment with Dox or peptide-Dox conjugate via tail vein injection. Data are presented as mean ± SD; n = 3; * p < 0.05, *** p < 0.001, **** p < 0.0001; Student t-test and one-way ANOVA were used with significance level (α) set at 0.05.

IHC Analysis of Cell Proliferation and Apoptotic Markers.

Figure 5 shows the IHC analysis of tumor proliferation and apoptosis biomarkers in tumor tissues following treatment of mice with saline, free Dox and peptide-Dox conjugate. In line with what was observed measuring tumor volume and luminescence measurements (Figure 3), at a dose of 2.5 mg/kg of Dox or Dox equivalent, only peptide-Dox was able to reduce the expression of PCNA and Ki-67 (indicating reduced proliferation) while increasing caspase-3 expression (indicating increased apoptosis). At this dose, the level of caspase-3, PCNA and Ki-67 expression in tumor tissue seemed similar for free Dox to that of saline treatment.

Figure 5.

Representative images showing IHC staining of caspase-3, PCNA, and Ki-67 in tumor sections from mice treated with saline (control), free-Dox, or peptide-Dox conjugate. Bar: 100 μm

Keratin 1 (K1) Expression in Mice Tissues.

To determine if there is overexpression of K1 in TNBC, we analyzed K1 protein levels in tumor tissues from mice and compared with normal mammary fat pad and liver tissues from mice. The IHC images show that K1 expression is upregulated in TNBC tumors compared to normal mammary fat pad and liver (Figure 6). K1-positive staining was significantly enriched in tumor tissues compared to the normal organs.

Figure 6.

Representative images of immunohistochemical staining for keratin 1. Tumor, normal mammary fat pad, and liver tissues from mice were collected, sectioned and slides were prepared. Slides were imaged using a rabbit monoclonal antibody against cytokeratin 1 (EPR17744/ab185628, abcam) at a magnification of 20x and 40x.

DISCUSSION

Anthracycline/taxane-based chemotherapy is still the preferred treatment option for TNBC patients.^19-20 We used Dox, an anthracycline drug, which is a moderately potent chemotherapeutic agent (IC₅₀ 0.2 – 1.5 μM), highly fluorescent molecule, and has decent water solubility.^13-14 Also, we continue to use Dox for investigating PDC therapeutic approach as it allows comparison between all previously synthesized PDCs in our studies.^13-14 We have synthesized four PDCs with Dox and peptide 18-4, where Dox is conjugated via ester, amide (in two PDCs) or hydrazone bond in the linker region.^13-14 This allowed a direct comparison of these PDCs with respect to the outcome. When incubated with human serum, the PDC degradation half-lives of the ester PDC and hydrazone PDC were 2 and 6 hours, respectively.^13-14 The two amide PDCs were more stable with PDC degradation half-lives of 18 h (Figure 1B, PDC used in the current study) and 48 h.^13-14 Further, it is observed that the amide PDC (Figure 1B) is more efficacious than the less stable hydrazone PDC. After 4-weekly injections, the tumor volume was significantly reduced for amide PDC-treated mice compared to the Dox-treated mice (Figure 3B-C), while in our previous study for the hydrazone PDC-treated mice the tumor volume was not significantly reduced compared to Dox treated mice at the same time point (after 4-weekly injections).¹⁷ Further, while Dox is a popular chemotherapeutic of choice for several cancers including TNBC, it is known for cardiotoxicity at high doses.^21-22 PDCs that deliver Dox specifically to tumor tissues and avoid heart and other healthy organs should decrease cardiotoxicity associated with this chemotherapeutic agent. Significantly lower levels of Dox (1.5 times lower) were observed in the heart of PDC-treated mice compared to free Dox-treated mice (Figure 4). Previously we showed that the uptake of a hydrazone PDC was significantly lower (1.4 times lower) in the heart for the PDC-treated mice compared to the Dox-treated mice.¹⁷

Dox was used as a control, not Doxil^®, because Doxil^® has a different pharmacokinetic (PK) profile compared to Dox.²³ Doxil^® is a pegylated liposomal formulation of Dox. The PK profile of the peptide 18-4 – Dox conjugate is expected to be more like Dox than Doxil^®, therefore, Doxil^® is not a good control for the PDC. We previously showed a circulation half-life of ~1.7 hours for another peptide-Dox conjugate (hydrazone PDC) in mice¹⁷ while Doxil^® is reported to have an elimination half-life of 20-30 hours.²³ In comparison, Dox displays a rapid decline of the initial plasma concentration (first distribution phase) with a half-life of 5-10 minutes, which is not observed for Doxil^®.²³

Peptide 18-4 is engineered starting from a lead 12-mer peptide p160 (VPWMEPAYQRFL).^{12, 24} Zhang et al. identified p160 for cancer cell targeting using random peptide phage display screening against WAC2 neuroblastoma cells.²⁵ The authors found that the p160 phage bound to several cancer cells (such as WAC2, Tet21N, and MDA-MB-435), and showed no binding to normal cells like 184A normal breast epithelial cells, HaCaT human epithelial cell line from adult skin, 293 human embryonic kidney cells, erythrocytes, lymphocytes, and monocytes. More importantly, the p160 phage was internalized by the cancer cells and the uptake mechanism was through receptor-mediated endocytosis. Starting with peptide p160 as a lead, we engineered peptides like linear 18-4 (10-mer) and a cyclic analogue for targeting of breast cancer cells.^{12, 16} The engineered peptides and the respective PDCs made with it are also internalized by cancer cells.^{12-13, 16} The involvement of K1 as a receptor for peptide 18-4 and its analogues in their uptake by breast cancer cells has been shown by us in a previous report.¹¹ Further, we showed that peptide 18-4 displayed negligible cytotoxicity toward cancer cells with cell viability of >98 ± 5% when incubated with the cells for 48 h at peptide concentrations up to100 μM.¹²

Keratins are well-established as diagnostic tumor markers for epithelial malignancies.²⁶ We found K1 as the target cell-surface receptor on breast cancer cells.¹¹ Although the structure and function of cytoplasmic K1 is known,^27-29 the cell-surface K1 is less studied and not much is known about it. Reports on other cell-surface keratins, like keratin 8, and a type III intermediate filament protein vimentin, show that these play an important role in cancer metastasis and other malignancies.^30-32 It is reported that cell-surface keratin 8 in cancer cells masks MHC I molecules on the surface of lymph node metastatic carcinoma cells and prevents their interaction with T-cell receptors (TCR) on the cytotoxic CD8+ T cells.³³ The MHC I – TCR interaction is an important mechanism by the cytotoxic T lymphocytes to eliminate malignant cells. Interference with this interaction is a strategy for immune escape of circulating tumor cells (CTCs), facilitating metastasis.^{30, 33} Cell-surface K1 may have similar roles, however these need to be yet explored.

Increased expression of cell-surface K1 in breast cancer and neuroblastoma cells over normal cells is reported.^{11, 34-35} Here, we show that tissues from mice carrying orthotopic human MDA-MB-231 (TNBC) tumors stain intensely for K1 while mice tissues from normal breast and liver do not stain for K1 (Figure 6). Other cancer types of epithelial origin that are likely to overexpress cell-surface K1 are under investigation. Identification of different cancer types overexpressing cell-surface K1 would broaden the scope of PDCs targeting cell-surface K1 for cancer treatment.

In line with higher expression of K1 in MDA-MB-231 xenografts compared to liver, we observed higher accumulation of Dox for conjugate treated mice compared to free Dox treated mice in tumor in our biodistribution studies (Figure 4). This has led to the potentiation of Dox activity in the inhibition of MDA-MB-231 tumor growth in mice compared to free drug at a low Dox dose of 2.5 mg/kg (Figures 3 and 5). We did not observe any change in the weight of animals in control and test treatment groups, pointing to the potential safety of the PDC under study at the administered dose.

In conclusion, our results suggest that K1 is a novel target for TNBC, and PDC (Figure 1) that consists of a linear peptide 18-4 targeting K1 is a promising novel modality for TNBC treatment.